In other posts I have sketched the Buddhist doctrine of 'No Self.' I now consider an early Buddhist argument for it. Here are the words of Buddha according to the Anattalakkhana Sutta, his second discourse, the Sermon on the Mark of Not-Self:
The body [rupa], monks, is not self. If the body were the self,
this body would not lend itself to dis-ease. It would be possible
(to say) with regard to the body, 'Let my body be thus. Let my body
be not thus.' But precisely because the body is not self, the body
lends itself to dis-ease. And it is not possible (to say) with
regard to the body, 'Let my body be thus. Let my body not be thus.'
Buddha then goes on to argue similarly with respect to the rest of the five aggregates or categories of personality-constituents (khandhas, Sanskrit: skandhas), namely, feeling (vedana), perception (sanna), consciousness (vinnana), and mental formations (sankharas). All are claimed to be not-self. Thus we are told that feeling afflicts us and is not amenable to our control, whence it is inferred that feeling is not one's self, not one's own inner substance. The tacit premise of this enthymematic argument is that one's self would have to be something over which one would have complete control. The tacit premise is that the self is something wholly active and spontaneous and self-regulating. And it is clear that something wholly active will not suffer: to suffer is precisely to be afflicted by something external over which one has no control. To suffer is to be passive. An agent in excelsis is an impassible agent. (In the West, impassibility became one of the divine attributes.)
After arguing that each of the personality-constituents is outside of our control and brings suffering, Buddha argues that each of the constituents is impermanent and for this reason as well is lacking in
self-nature. The over-all argument of the Anattalakkhana Sutta may therefore be reconstructed in a generalized form as follows:
1. If anything were the self, then it would have two properties: it
would not be liable to disease, decay, destruction or change
generally, and it would be self-determining, i.e., it would have
complete control over itself.
2. But nothing in our experience has either of these two
properties, not the body, or feeling, or perception, or
consciousness, etc.
Therefore
3. Nothing in our experience can be identified as the self.
I have structured the argument so that it is not only valid in point of logical form, but also has plausible premises. (A charitable reconstruction can aim at nothing less.) Thus the addition of 'in our experience' in premise (2) makes this premise more plausible than without the addition. But this gain in plausibility exacts a price: the conclusion (3) cannot then amount to the unrestricted anatta doctrine according to which nothing at all is a self or has self-nature. It is one thing to say that nothing in our experience can be identified (veridically) as a self, and quite another to say that there is no self.
The Pali Buddhist, of course, wants to arrive at the conclusion that there is no self at all, that nothing at all has self-nature. I would insist, however, that one cannot validly move directly from (3) to
4. Nothing can be identified as the self: there is no self at all.
For it may well be that the existence of a (transcendental) self that is not experienced is a necessary condition of establishing that whatever is experienced is not a self. After all, if I come to see that my body, feelings, perceptions, and so on cannot be identified as my very self, then it is presumably I myself who come to this insight. If I come to reject body-identification, feeling- identification, etc. as false self-identifications, then presumably there must be at least the possibility of a true self-identification, even if only of the tautological form, 'I am I.' If every self-identification were false, then 'I am I' would be false. But that is either a contradiction, or implies that there is nothing that body, feelings, etc. are distinct from, which is again incoherent. For if 'I am not this body' is true, then in some sense I must exist as that from which my body is distinct. The first-person singular pronoun cannot be wholly referenceless if it is to be true, as it is true, that I am not my body, my feelings, my thoughts, the pain in my neck, etc.
Furthermore, if anyone needs and desires liberation, it is presumably in every case I myself who needs and desires it, and I myself who, if all goes well, achieves it, and indeed achieves it on the basis of my own insight into my non-identity with any of the five khandhas or with the psychophysical complex composed of them. The self who needs, desires, and possibly attains liberation is obviously distinct from each of the khandhas and from the psychophysical complex. My body, obviously enough, cannot come to realize its non-identity with itself, for the simple reason that it is not distinct from itself. The same holds for each of the khandhas, and for the lot of them taken together.
And to suppose that no one desires, needs, or attains liberation would appear to make hash of the whole Buddhist system of soteriology. Buddhism is an existentialist system in roughly Kierkegaard's sense: it is the salvation of the "existing individual" that is the unum necessarium and sole desideratum. It is therefore arguable that the existence of a transcendental self is a cognitive and soteriological presupposition: it is presupposed if there is to be the insight that no object of experience is veridically identifiable as one's very self, and it is presupposed if there is to be something that is saved from the samsaric predicament.
The gist of the control argument is this. There is no evidence of a self since nothing with which we are acquainted is immutable, and nothing with which we are acquainted is something over which we have complete control.
But this raises an obvious question: Isn't the standard for selfhood being set unattainably high? The argument isd tantamount tosaying that if I am not God or a god, then I am not a self. Arguably, God to be God must be impassible; but must a self to be a self be impassible?
For a richer and more rigorous development of this theme, see W. F. Vallicella, "No Self? A Look at a Buddhist Argument," International Philosophical Quarterly, vol. 42, no. 4 (December 2002), pp. 453-466
Recent Comments